The Great Lightbulb Conspiracy (2014) | Hacker News Reader

The Great Lightbulb Conspiracy (2014)

spectrum.ieee.org

128 pointsmmastrac85 comments

Removing light bulbs from retail market and enforcing mercury based CFLs - that was the latest act of such a cartel. Good the the paradigm shift to LED was quicker and different then they hoped. Now Philips, GE and OSRAM are a shadow of its former self, and little players in the LED industry. And Asian build LED are of a often of a higher quality and last longer. (well probably all LEDs are built there anyway, even the one from OSRAM that whats to sell that part of the company) And Asian built traditional light bulb are available and last longer than the older cartel build bulbs.

> And Asian build LED are of a often of a higher quality and last longer.

I'd like to believe that. I got burned by so many cheap Chinese LEDs (quite literally, too - some common failure modes of LED bulbs have them getting almost as hot as incandescent ones) that now I'm willing to pay premium to be sure the bulb won't turn into disco stroboscope after two days of usage.

Planned obsolescence would be changing the base on the lamp every few years so that people would have to buy new fixtures when they can't buy replacement bulbs anymore. Author is not using that term correctly.

The actually interesting story is about the market failure where customers have no choice but to prefer the thing they can measure (bulb life vs cost) over the more important thing they can't measure (electricity cost over life of bulb).

> Planned obsolescence would be changing the base on the lamp every few years so that people would have to buy new fixtures when they can't buy replacement bulbs anymore.

Have you bought a light fixture recently? That sounds like the new state of affairs. Aided by government regulations that require new lamps not be able to accept older "inefficient" bulbs, there's been an explosion in the variety of sockets.

I've converted every lamp in my apartment to LED except for one, because it uses a GU24 base that is very difficult to find in anything other than CFL.

Which government? I have not seen this in the US, and I purchased quite a few lamps in the last couple of years.

If it happens here I'll be extremely sad. CFLs and LEDs have an extremely strong negative effect on mood for me. Fortunately I can still buy halogens, which aren't quite as cozy but suffice.

Sounds like you react to bad LEDs. Good ones have stable power supplies that don't ripple, and phosphor coatings that produce fairly even frequency distributions. LEDs for room lighting are still relatively immature, but they're getting there quite quickly.

Planned obsolescence or built-in obsolescence in industrial design and economics is a policy of planning or designing a product with an artificially limited useful life, so it will become obsolete, that is, unfashionable or no longer functional after a certain period of time.

- Wikipedia

Seems fitting to me.

The longest running light bulb is in Livermore, CA and has been running since 1901. It's installed and maintained in a fire station.

https://en.wikipedia.org/wiki/Centennial_Light

you made my night!

Obviously an environmental disaster if everyone had these.

But still, a quaint thing.

From the Wikipedia article

> The Centennial Light was originally a 30-watt or 60-watt bulb

Why? Everyone, where I'm from anyway, now has CFLs and LEDs, mostly, and the environmental disaster rolls on.

Don't forget though that there really is a direct correlation between efficiency, brightness, and lifetime. Your electric stove makes a great long lasting light. Just turn all 4 burns on full. It will last virtually forever.

Yes it's not such an immoral decision as it is presented. Light bulbs that have energy consumption in the KW range because people want them to last forever would not be good.

What I would like to know is, how did Thomas Pynchon know about this when he wrote Gravity's Rainbow in the early 1970's? Research pre-internet.

http://www.theparisreview.org/blog/2014/10/07/planned-obsole...

> Of course, given the collective ingenuity of the cartel’s engineers and scientists, it should have been possible to design a lightbulb that was both bright and long-lived.

And not energy efficient.

The article lightly touches on it, but the physical reality is that the better the electrical efficiency the less time the bulb lasts.

It's a tradeoff that has not changed in 100 years.

So yes, the cartel did mandate 1,000 hours - but they also saved untold amounts of electricity over then next 100 years.

This is the one big thing missing in most reports about the lightbulb lifetime. Even consumers profited via power-savings from this lifetime agreement.

Seems like an important point, but I think it would also be worth to see a comparison of energy saved by end users through increased efficiency, and energy used on additional manufacturing, distribution and sales caused by the need to replace bulbs more often.

This is a general point about all "eco" technologies too - a device is not "eco-friendly" if it's total manufacturing and distribution energy costs are higher than the "eco-unfriendly" alternative.

Conveniently, the retail price includes the direct expenditures on energy to make and transport the item to the store.

I recall the electricity rate being around 0.08 USD/kWh around here a while back, so 1000h of 100W consumption would cost $8. I don't remember exactly what 100W plain light bulbs cost, but it certainly was a lot less than that.

I'm talking about energy costs and environmental damage, not the dollar price. Contrary to popular opinion, the retail price rarely if ever captures the environmental impact of a product.

I can't believe an LED that adapts the colour temperature to the time of the day like F.lux does isn't a standard yet.

LEDs are great but boy are they agressive to the eyes.

(Yes that's a bit unrelated, I figured I'd rant and hope some poeple starts mass proucing it thanks to my message)

I have no trouble getting hold of 2700K LED bulbs here in Australia.

I agree though, the cold white 4500K or higher bulbs are really only suitable during the day. Every bulb in my house, CFL or LED, is 2700K.

The difference between a 2700k and 3000k bulb can be quite large, too. I prefer the ones I've gotten labelled 2700k, 40w equivalent.

Better LED bulbs also have better power supplies, and therefore less ripple.

I just ordered a some from Osram Lightify that you can control the colour temperature and brightness. I'm going to set them up with Home Assistant to do exactly this, and at €20/bulb (for GU10) they aren't too bad (compared to the coloured Philips Hue which are €60/bulb). However as you say, it would be great if the bulbs would just do this themselves.

I changed my apartment recently and replaced all bulbs with Hue. Thanks for another idea to my list of hacks I'm trying to implement on them. Hue has a nice RESTful API that works over local network - I so far managed to get Tasker on Android to talk into it, and I'm moving to build a more advanced control centre on a Raspberry Pi.

Are LEDs aggressive to the eyes even if you manage to get them to have the same colour as ordinary light bulbs?

You mean spectrum - a big difference. "Color" means little or nothing. Very different spectra can have the same "color".

I highly recommend reading through this, very educational about how complicated "vision" really is: http://webvision.med.utah.edu/book/part-vii-color-vision/col... It's a long read but well worth it.

"Color" is something your brain creates, it's not "physical". Frequencies of light of course are a (the) major input, but it's not in any way a straightforward 1:1 one.

I mean colour, not spectrum. I don't expect, or at least don't count on, LEDs to reproduce a black-body spectrum.

I recognise that two light sources with the same colour, but different spectra, might give different results when reflected off things. Is that what you had in mind?

I think he meant that 2 things having the same colour, but different spectrum, can have different effects on your vision (and body?). The example that most easily comes to my mind is a LED screen representing a yellow colour, versus a lemon under the sun.

Shame he didn't elaborate, it seemed to be an interesting topic.

Every light source produces a spectrum - even a laser, albeit only small one.

Have a look at this link: http://www.soundandvision.com/content/led-vs-cfl-bulbs-color...

As you can see a lot of LEDs have quite a lot of blue. Which is exactly what is responsible for synchronizing your own "clock": http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2831986/

    > Not only does light reset the human circadian rhythm, but the same blue
    > light that has the strongest impact on dinoflagellates has equal power
    > to reset our own clocks — although most visible wavelengths can reset
    > the clock, the blues do the job with the greatest efficiency.

Which is the whole reason f.lux became so popular.

Also brightness. At night you just need a bit of light so you can still see stuff, without the light disturbing your circadian rhythm. At daytime you want the light to be much much brighter.

Related: I have noted over the years that incandescents with dimmer switches basically never burn out, and those with normal switches tend to break during the first second of being switched on. That's led me to the following thought: would it be possible to build a bulb with a built-in fade-in, either via bandpass/lowpass filtering or some sort of fading mechanism?

I promise you that bulbs on dimmer switches do burn out. I just bought a house and the kitchen light on a dimmer switch burned out within a week of my moving in. No idea about the age of the bulb, so it's not very good data, but... Sadly, it's apparently still a crapshoot whether any given LED bulb works without flicker on any given dimmer of unknown design; I have only found those with flicker so far.

I have heard that such devices already exist. They are a NTC cap that goes over the end of the bulb.

http://donklipstein.com/bulb1.html#ss

For 90 years, lightbulbs were designed to burn out. Now that's coming to LEDs

https://news.ycombinator.com/item?id=12103244

That article had no evidence on its face, and was debunked in the comments.

No they weren't and no it isn't.

I miss the times when every light fitting was the same (one size, bayonet style). I don't know how many different types and sizes of fittng we have. It's a pain replacing bulbs now. I preferred the light from the old incandescent bulbs too and if they give off a little heat, well living in the temperate zone we need a little heat after dark.

The light is better, but the heat is usually not where you want it.

Won't it amount to the same thing as a heat source anywhere else in a room, with convection and radiation?

Modern heat pumps are three to four times more efficient in their use of electric power than simple electrical resistance heaters.[1]

I have a 7.1kW heat pump, it uses a little over 1.79kW of power generate that heat, so a thermal efficiency of 3.96:1. Any light bulb, whether it's incandescent or CFL or LED produces heat a 1:1 thermal efficiency.

1. https://en.wikipedia.org/wiki/Heat_pump

Do many people have heat pumps? Most people I know in the Midwest make 100% if heat from burning natural gas. So in the winter not a huge loss (just the delta on cost between electricity and natural gas). In the summer obviously you pay double..

Heat pumps / reverse cycle air conditioners are very popular in Australia, but the temperature here is fairly moderate, even in Tasmania where I live there are few places where it's below zero for more than the over night period. Heat pumps don't work so well sub zero.

Many of the old incandescent bulbs have been replaced with halogen bulbs with filters so they produce similar light. But they're so much more efficient that I can no longer rely on them to warm a room at night during the fall. Also they burn out in less than a year.

Not altogether different from tech. today. Ie. Year 1 nobody owns a light bulb, so sell an incredibly efficient bulb. Year 10 everybody owns bulbs so cheapen the product to increase sales. While today we have incredible free service to grow market, then when the market had matured and growth falls and lock-in is complete reduce service to monetize.

Somethings never change.

The article mentions CFLs and LED lamps, which are definitely subject to the same engineering pressures:

This would result in increases of candlepower of 11 and 16 percent respectively.” That boost in illumination, he suggested, “would be acceptable to all flashlight users” despite the fact that the higher current would shorten not just the bulb’s life but also the battery’s.

The cartel’s justification for these changes was that at the higher current levels, the bulbs produced more lumens per watt. Alas, more current means not only more brightness but also higher filament temperature and therefore shorter life.

I have no doubt that LED lamps could be engineered to last essentially indefinitely --- well past the 100+ year record of early incandescents[1] --- but most consumers are lead by the advertising to seek higher brightnesses, which are achieved by running the LEDs at life-shortening higher currents and temperatures.

Newer ICs with small feature-sizes are subject to far more current density and electromigration stress, to the point that CPUs are not expected to last more than a few years, whereas older ICs may remain functional for many decades.

Memory and storage densities are also subject to the same effect, where it seems "capacity and speed is everything" and data retention and endurance have been sacrificed. NAND flash started out at 100K+ cycles per bit and 10+ years retention; now it's more like 1K and 1 year.

[1] https://en.wikipedia.org/wiki/Longest-lasting_light_bulbs

Honestly, I'd be fine with optimizing products like that if two things were true:

- close-to-perfect recycling infrastructure; i.e. all the stuff that breaks down and is thrown away actually gets reused to create a new generation of stuff, with minimum additional input of energy and material

- civilization was a stable thing; making things throwaway depends on ensuring there's always a replacement to buy - which will be sort-of the case only until the world goes to war again (and/or western economy goes seriously belly-up), at which point we (those of us who survive, that is) will be cursing previous generations for making all advanced technology break down after a year of use, often on purpose

Alas, those things are not true, and so I despair over the wastefulness of what "the market decided".

> Honestly, I'd be fine with optimizing products like that if two things were true:

If you're talking about technology being expected to break down and be cheaply replaced on the scale of single-digit years, it sounds like a never-ending nightmare. I'd need to add two more points, at least.

- Strict upgrades (that is, there must be nothing about a new version of a product that's worse than the one before it); if one feature/technology was present in the older version of something, the new version must provide the same capability. I'm thinking of the old hardware I keep around because of "That one thing that only runs in Windows 98" and "That one thing I can only flash over a parallel port".

- A mechanism to automatically, securely, and privately image the new machine with the previous machine's configuration and data. The nests I build on my machines feel like home.

Given only your first two points, I still wouldn't buy a machine that I didn't expect to last at least 5 years, and preferably 10.

I was thinking primarily about appliances - lightbulbs (dumb or smart), washing machines, cars, cellphones / smartphones, etc. Those two points you added don't apply to most of them.

That said, I'm not sure if I agree with the first one. At least not completely. I'd like to have some ability to recover old/special features but I wouldn't mind if it required some extra work (like tracking down that motherboard model that still has parallel port on it). In the reality where all technology is being obsoleted fast and my two points hold, whatever you want to program over parallel has probably broke down already anyway.

As for the second one - we don't have anything like that even today in wide use, and we cope :). On Unix you end up moving your dotfiles; on Windows I used to spend an hour or three for setting things up "my way". Personally I don't expect a new machine to immediately feel like the old one - but I do expect to be able to make it so.

I guess that my overall point is that I dislike almost everything about the idea of disposable possessions, and I would resent being forced into it. I could imagine leasing a car, my washing machine, or another appliance if the TCO was reduced, compared to outright ownership for 10 years, but nothing that I actually care about.

> I'd like to have some ability to recover old/special features but I wouldn't mind if it required some extra work (like tracking down that motherboard model that still has parallel port on it). In the reality where all technology is being obsoleted fast and my two points hold, whatever you want to program over parallel has probably broke down already anyway.

If they're willing to replace the parallel device with a functionally-equivalent piece of hardware that connects to a bus that the new hardware provides, then that would be equivalent. I'd just want to avoid situations like this: If I took pains to keep OtherOS on my PS3, the replacement hardware must not have a firmware that's incompatible with my use of it. If I've got a gaggle of microcontrollers and a parallel-port programmer for them, the new hardware must either let me program them.

> On Unix you end up moving your dotfiles; on Windows I used to spend an hour or three for setting things up "my way".

On Unix, I move my dot files, a couple hundred gb of data, install about 500 packages, and end up tracking down, compiling, and installing a few things that aren't usually included in my repos. In Windows, a lot of the data was already transferred from the Linux side ("Unix" is invariably "Linux", on my machines), and I spend the next week tracking down installers for software, re-downloading games, etc. For the next 6 months, I'll be running into things that I need to track down and re-install. I cope with this by keeping hardware until it's absolutely unusable for what I want to do. If someone expected me to upgrade more frequently through a forced process, I'd expect them to ease my way as much as humanly possible.

Not saying you're speeding FUD, but do you have any sources discussing CPUs (or GPUs) lasting only a few years? In my experience, desktop, server, and mobile CPUs work fine after 10 years, and I'd expect the same of current tech.

No FUD, just physics. The parent comment mentioned electromigration[1].

From the Wikipedia article:

Electromigration is the transport of material caused by the gradual movement of the ions in a conductor due to the momentum transfer between conducting electrons and diffusing metal atoms. The effect is important in applications where high direct current densities are used, such as in microelectronics and related structures. As the structure size in electronics such as integrated circuits (ICs) decreases, the practical significance of this effect increases.

It remains to be seen how long modern CPUs (ICs in general) last. My i5 laptop is still running at 3 years, so there's that.

1. https://en.wikipedia.org/wiki/Electromigration

The counter-argument to our fear of planned obsolescence here is the ridiculous speed of technological improvement in LEDs relative to incandescent lighting. LED's cost per lumen falls by a factor of 10 every decade (see Haitz's Law), while a user's actual lighting needs may never exceed 800 lumens for an standard A19 bulb. Any attempt to form a cartel in this market would be swiftly defeated by ever-falling unit economics. Far from a cartel, we should let the government create a standardized LED lifespan rating based on the actual needs of consumers. For example, if we know the average renovation cycle for single family houses in the US is ten years, then we should make LED bulbs that are rated for 10 years. This way, manufacturers can deliver these products at at the most affordable price point, and consumers for the most part won't have to worry about replacing their lights ever again.

LEDs suck less, and are much better than garbage CFL, but are still more expensive and less versatile than the old bulbs, particularly when you need a bright bulb.

Fortunately, you only need a bright bulb for flashlights and for fixtures that are intended for incandescent bulbs. For general interior lighting, a large array/string of relatively dim LEDs is usually fine.

Not sure why you would think anyone today would use an incandescent bulb in a flashlight. This is where LEDs just took over. I have a pencil sizes LED flashlight that matches the brightness of my Maglite only when I put it in power-saving mode - otherwise its about 10x as bright.

except that many LEDs flicker horribly because they have bad electronics, and the consumer has no way to judge quality.

This really is the crux, unlike fluorescent (including CFL) and traditional incandescent lamps LED's are a lot more sensitive to having consistent DC power, a crappy little capacitive dropper is not appropriate to power them, yet a lot of manufacturers decided to cheap out on an additional diode to build a proper bridge rectifier. The amount of electronics to properly drive an chain of LED's is so simple and cheap I really can't figure out why they are shaving tenths a penny or two (considering the bulk rates they can get these parts at) on lamps that are sold at pretty high profit margins.

I really wish there was someone like bigclive that just tore down LED lamps sold in the US and posted reviews on Amazon, etc. about the internal driver circuitry.

Your comment captures the problem far better.

For 100 years, if a bulb burned out or you bought a lamp, you went a bought a sodding lightbulb. The choices were clear... good brand/shit brand. 20/40/60/75/100/150/3way. Soft white or not soft white. They cost like a quarter.

Now the Home Depot lightbulb aisle is as complicated as the screw and fastener aisle. My guess is the average bigbox home store has 50 SKUs on the shelf for 60 watt equivalent bulbs. It's ridiculous.

Go to Ikea.

They have about four bulbs for each socket (pear and candle, dim and bright), and the ones I bought a year ago are very good.

I think they have more incentive to sell good quality bulbs than the average store, since they sell the fitting and market being environmentally friendly.

Of the ones I bought, some were made in China and some in Germany.

There are some US states whithout an IKEA, such as Maine.

Incidentally, in Maine, incandescent light bulbs are almost 100% efficient for much of the year as they produce both light and heat.

When using electricity for heating, a heat pump is more than 100% efficient.

I live in Denmark, where regulations mean new buildings can't have simple electric heaters installed.

Yep, modern heat pumps are up to and above 4:1 thermal efficiency. References in my other comment.

100% is a bit of a deceptive claim, since 1) we can exceed 100% with heat pumps, and 2) pretty much _every_ electrical appliance in an enclosed space is 100% efficient at converting energy to heat. 3) For the most part, the electricity being used to power those incandescents is produced with 25-50% efficiency. Burning gas to make heat to run a turbine to produce electricity to transmit it to your home to heat a coil to make light is not very efficient. Ideally it's combined-cycle (waste heat boils water to spin another turbine) but it's still not as good as just burning that gas in your furnace.

My apartment, lovely as it is, has vile, ancient, horrible resistive electric storage heaters. I live in a somewhat cold climate. I suck it up and pay for renewably-sourced electricity to assuage the guilt because there's no way I'm going to win in a fight to install heat pumps or gas-fired heating with my landlord.

In this limited scenario, incandescents make sense. If I owned this place those hunk-of-crap heaters would be in the skip tomorrow and all the incandescent bulbs with them.

Electric heating is a huge red flag - living in the northeast US, that tends to indicate that whoever built the building is not from around here, and perhaps came up from the South, or some other area that does not have real winters. Usually where you find electric heaters, you'll also find wildly insufficient insulation, foundations that aren't up to frost, and any number of other design flaws that bite you in the ass when it is below freezing for four or five months out of the year.

Natural gas or heating oil is far more economical, although nothing beats the warmth provided by a good wood stove.

I agree, though in this case it has more to do with Dublin apartments generally having crap heating facilities. The landlord knows full well the costs he imposes.

Of course, I could also have looked for an apartment with gas heating, but the reality is the rent here is low enough that simply offsetting the environmental cost of electric heating with power sourced from wind (which is where almost all of this island's renewable energy comes from) is still a better deal.

It may be worth noting that it gets cool, but not especially cold, here, at least by New England standards. In January the average low is 3C, so still above freezing, and snow is rare, though it does happen.

I really wish there was someone like bigclive that just tore down LED lamps sold in the US and posted reviews

There is such a guy "electronupdate", but I think he's from north of the border. Close enough. His older teardowns were better, because the bulbs themselves were better. Now there's a race to the bottom (of course).

If you watch his teardowns you will see that the older Philips bulbs are superior to all others.

E.g. here's a very good review: https://www.youtube.com/watch?v=OWKAub54sLU

He's also done teardowns of some real crap bulbs.

There are also some "bigclive" LED bulb reviews on Youtube but I haven't watched them.

"Consumers for the most part" do not regularly renovate their houses. Most of them are renters. And paying someone else to change your lightbulbs doesn't protect you from having to get them changed.

This. I know that the "10 years" was just an example, and I don't have the relevant data on home remodels, but 10 years on average in the US seems ridiculously fast to me. I would guess that the average is more like 30-40 for those doing renovations, and like parent says, many aren't renovating.

So instead of a private cartel regulating light bulb lifespan and fining manufacturers whose bulbs last more or less than that time, you want the government to do the same?

There is a pretty big difference between a mandatory lifetime and a minimum lifetime.

As far as I know, minimum fitness-for-purpose standards are pretty good for the consumer.

That's a fair criticism. I did conflate two different things there, and I was needlessly snarky. (And I upvoted your comment for pointing it out.)

I also failed to address the thing that bothered me in the comment I replied to:

> Far from a cartel, we should let the government create a standardized LED lifespan rating based on the actual needs of consumers. For example, if we know the average renovation cycle for single family houses in the US is ten years, then we should make LED bulbs that are rated for 10 years.

Why should it be the government's business to determine what "the actual needs of consumers" are? How does the government know what the average renovation cycle for single family homes is? What about people who want to spend more for longer-lasting bulbs, or spend less for shorter-lived ones? What about other types of homes that don't fit the government-specified "average renovation cycle"?

How do companies innovate when the government has said, "This is the standard specification for an LED light bulb. You will make them to our standard or you won't make them at all."

I just find the very concept offensive that this should be any business of the government. Life is full of trade-offs, and it seems to me that the government - the instrument of force - is least capable of anyone to make these trade-offs.

Regardless of my personal feelings, there is a real cost in getting the government involved in these things. Government intervention never comes for free.

> Why should it be the government's business to determine what "the actual needs of consumers" are?

The government has signed up to various international treaties around ewaste and other eco indicators.

Consumers only buy on price.

Government sets minimum standards, which forces manufacturers to build their minimum cost devices to that standard. Manfs can also offer better devices for more money.

> How do companies innovate when the government has said, "This is the standard specification for an LED light bulb. You will make them to our standard or you won't make them at all."

Your mobile phone is intensely regulated across a number of different regulators. There seems to be plenty of innovation in cellphones.

Thanks for your comment, you raise some good points. I do have to question a couple of things:

> Consumers only buy on price.

Really? Everyone I know also looks at reviews, brand reputation, and other factors. Of course you're right that sometimes we just buy the cheapest option. :-)

> Your mobile phone is intensely regulated across a number of different regulators. There seems to be plenty of innovation in cellphones.

Is the innovation in the areas governments regulate, or in the areas they don't? I just made a list of 20 or so key attributes of a cellphone, and I don't think they are things that are regulated. (I was going to post the list, but had second thoughts that it may seem argumentative, and they seem pretty obvious anyway.)

> Really? Everyone I know also looks at reviews, brand reputation, and other factors. Of course you're right that sometimes we just buy the cheapest option. :-)

Anecdata time! :).

I think that this sometimes turns out to be more often than not :). Not many people have time to read reviews and check brand reputations for everything they buy. Personally, I divide shopping decisions into two groups - where I care about quality, and where I don't. In the former group, which includes mostly expensive stuff, I'll check some reviews. In the latter, I don't care. The latter group mostly consists of things that cost about as much or less than my time spent on doing research.

But from what I've seen (and experienced), a lot of things move from the "check reviews" to "buy on price" group the moment one is low on cash and/or low on free time. Anecdata - my mother used to buy CFLs but at some point switched back to incandescents (the ones that are illegal in EU, but are being sold in every corner shop anyway, just labeled as "not for household use") just because she didn't have time / energy to go to the big store that sells decent CFLs at a reasonable price. I only upgraded her apartment to LEDs after few of those cheap-ass incandescents literally exploded, sending glass shards all over the kitchen.

> I was going to post the list, but had second thoughts that it may seem argumentative, and they seem pretty obvious anyway.

Please post that list, I'm interested. Even if it seems obvious, it may not be obvious to everyone.

I think your anecdata is right on the money!

(Pun unintended, but now that it's there I'll take credit for it...)

Yeah, when I said "everyone I know looks at reviews" I meant "at least some of the time". :-)

So my list of cellphone characteristics I didn't post - I was thinking of things like this:

Size, shape, and weight. Battery technology and capacity - removable or not? Display technology, size and resolution. Resistive or capacitive touch? Stylus? Physical keyboard? Processor technology, how many cores, what instruction set, what GPU? Memory size? Storage size? External storage (SD)? Charger standard? External audio and digital ports? What kinds of sensors? Operating system and application software. What programming languages are used. How software is distributed. Who is allowed to write software.

I don't think many of those are specified by a government agency. At least I hope they aren't!

I personally like CFLs in the right context, but if you want instant light, dimming or cooler light, they aren't great. https://www.ovoenergy.com/guides/energy-guides/energy-saving...

Great eponymous documentary about this and planned obsolescence http://youtu.be/-1j0XDGIsUg

Blackbody radiators and millions of years of evolution are hard to beat. I like the energy efficiency of CFL and LED but I do not like the spectral power distribution or the decay rate and ensuing hue shift of these light sources.

And it's not something only experts notice. If you have multiple bulbs in the same proximity, one fails, whether replaced under warranty or you just buy a new one with identical color temperature specs, it'll look different from the others.

Recapturing heat loss from incandescent to make them more efficient, maybe even on the order of today's LEDs seems more promising to me than arresting decay rates and/or the hue shift problems.

> heat loss from incandescent to make them more efficient, maybe even on the order of today's LEDs seems more promising to me

This is already happening - you can apply a 'hot mirror' coating to reflect infra-red radiation back into the bulb, while allowing visible light out. This is a very old idea - here's a patent from the early 1960s which states that the basic principle was already in use:

http://www.google.com/patents/US3174067

And another which applied it specifically to coating a halogen incandescent light bulb:

http://www.google.com/patents/US3209188

Here's a report / puff piece from 2013 asserting that this technology is currently taking off and claiming that "by 2018 the efficiency of a hybrid halogen incandescent light bulb will exceed that of the current compact fluorescent lamps (CFL) and equal light-emitting diodes (LEDs) on the market by 2023":

http://www.smithmillermoore.com/Pdfs/technicalarticles/3-6-1...

Wow, thanks for posting this and to the investigative journalist.

Now I know whom to blame for all the heartburn that I (and family) had from the continuously fusing bulbs.

Milton Friedman famously claimed monopolies/cartels would not exist without government assistance and that there are only 2 examples of private monopolies that lasted long period of time (De Beers and NYSE). I suppose he was very wrong. https://www.youtube.com/watch?v=tdLBzfFGFQU

The article suggests Friedman was right - this cartel was very short lived.

"Powerful and influential though it was, the Phoebus cartel was short-lived. Within six years of its formation, the cartel was already starting to struggle."

Friedman was also right that the cartel existed primarily due to government assistance:

"GE’s licensing of its basic lightbulb patents gave rise to yet more alliances, most notably the powerful Patentgemeinschaft (“patent pool”)..."

how many people does it take to screw in an everlasting lightbulb?

0. Because it either don't exist or already screwed in by someone else years ago.

Another neat side effect, if it's everlasting, it will probably not be mounted in a changeable fixture.

Reply on news.ycombinator.com